This competing continuation application grows out of a parent project to study how the structure and organization of drug use social networks affect HIV transmission. The parent project successfully recruited 169 networks (126 drug user networks and 43 nonuser networks). These networks were described by 300 respondents and contained over 2,000 individuals. Discoveries from the parent project showed that many long-term relationships were often characterized by short periods of intense contact separated by periods of non-contact. It was also found that emotional closeness appeared to be the best predictor of network stability. Mathematical models of the HIV epidemic suggest that elements of network change may have significant impact on the spread of HIV. Concurrency (multiple partners at the same time) and turnover (changes in network membership over time) are both implicated in high rates of HIV transmission. The purpose of this competing continuation application is to investigate the role of these two factors on HIV transmission. The continuation project proposes to collect a sample of 175 persons, both drug users and nonusers, who participated in the parent project to be interviewed at 3-month intervals for 4 years about their drug use, sex, and other network relationships. Data will be collected at the individual level (participant's demographic characteristics; economic, drug use, and sexual behaviors; and psychosocial functioning) and relationship level (joint injection, drug use and sex behaviors, and emotional closeness). In addition, data will be collected on all risk partners (injection and sex partners) as well as drug user partners, economic partners and emotional closeness partners. Innovations in relationship data collection based on experiences in the parent grant will improve the accuracy of relationship data. Data at each wave will be compared with data from previous waves, and changes will be explored quantitatively and qualitatively. The existing data and the proposed data will provide a unique opportunity to investigate the effects of network turnover and emotional attachment on HIV transmission risk.